Building Supply Chain Resilience: How SMBs Navigate Tariff Disruptions with AI-Driven Planning

Tariffs and supply chain disruptions are no longer occasional shocks. For many small and mid-sized businesses, they have become part of day-to-day planning. Costs shift quickly, lead times stretch, and supplier reliability can change without warning. In that environment, building supply chain resilience is moving from a nice-to-have to a practical way to protect service levels, manage cash, and keep growth on track.

In a conference session on navigating tariffs and disruptions, Jefferson Barr, SVP of Global Marketing at Netstock, shared what he is seeing across thousands of SMBs and what their planning data reveals about real-world adaptation. The core message was pragmatic: SMBs do not have the same buffers as large enterprises, so resilience often comes down to visibility, faster decisions, and smarter inventory optimization. The session focused on how organizations are responding to tariff pressures, how inventory strategies are shifting, and why AI-driven demand forecasting is gaining traction as a decision-support layer that reduces manual effort and highlights next steps. The result is a clearer picture of what is working now and what supply chain teams should prioritize as we head into another year of uncertainty.

Tariff Disruption is Forcing a New Baseline for Supply Chain Planning

Tariffs remain the elephant in the room because they change the economics of purchasing overnight. Barr framed the current environment as one in which organizations are making planning decisions while policy and pricing remain uncertain. The impact is evident immediately in inventory costs, purchasing behavior, and supplier negotiations.

Most SMBs Feel the Impact, But They Respond Differently

Barr referenced benchmark findings indicating that 63 percent of SMBs reported tariff impacts, either through higher costs or reduced purchasing. Two common reactions emerged:

Absorb higher costs to protect inventory and customer service

Cut purchases and accept higher risk to service levels

Nearly half of the respondents absorbed inventory maintenance costs, effectively prioritizing customer service over margin. A smaller portion cut can reduce cost exposure but increases the risk of stockouts and dissatisfaction.

Supplier Pricing is Volatile, and Many Lack a Plan

Tariff volatility makes supplier rate strategies more important, but the data suggest many organizations still lack a clear approach. Barr noted that 49 percent were not able to lock in supplier rates, leaving them exposed to volatility. Long-term contracts appeared more common than financial hedging, yet a significant share reported no strategy at all, and that figure increased year over year.

For planning teams, this is a critical signal. When supplier rates are unstable and contracts are uncertain, forecasting accuracy becomes less valuable if replenishment decisions cannot adapt quickly. The planning system must model uncertainty and guide trade-offs among cost, availability, and service-level management.

Freight and Sourcing Choices Point to Adaptability, Not a Clean Shift

The discussion also touched on shifts in inbound modes and sourcing strategies. Sea freight usage declined while land freight increased, and many companies split modes. This split suggests organizations are actively experimenting with options, balancing speed, cost, and risk rather than committing to a single approach.

Supplier sourcing showed a similar pattern. The most common strategy was split sourcing across domestic and offshore suppliers. Offshore preference remained significant, driven by cost and availability. Domestic preference existed but remained a minority.

The takeaway was not that nearshoring is a solved strategy. It is a trend that many want, but few can execute quickly.

Inventory Optimization is Evolving: Strategic Buffers, Dead Stock Risk, and Service Level Pressure

When tariffs and disruptions raise uncertainty, inventory becomes both a safety net and a financial constraint. Barr emphasized that SMBs face a delicate balancing act: they need service levels to stay competitive, but they often lack the working capital to carry excess inventory for long.

Strategic Excess Stock is Becoming a Deliberate Hedge

A key insight from the benchmark data was that some excess inventory is intentional. Barr reported that 30 percent of SMBs said over 30 percent of their excess stock is strategic, used as a buffer against risk. This is a notable reframing:

He anchored this in a familiar pattern: a small portion of SKUs typically drives a large share of revenue, so teams prioritize availability for those items even if it means carrying more inventory than usual.

Dead Stock is Increasing for Some, and It Hits Cash Immediately

At the same time, the risk is that hedging becomes overbuying. Barr highlighted that a meaningful share of firms carry dead stock, and that the mix is shifting: firms with low dead stock declined, while firms with high dead stock increased. His explanation was straightforward. Many organizations purchased more inventory during periods of uncertainty, and some of that inventory is now sitting idle.

He shared an image from a finance leader that captures the pain: a warehouse worker walking through the facility and seeing pallets of cash on racks. Dead stock ties up capital, occupies space, and can erode margins later through discounting.

Most Organizations Rely on Promotions to Reduce Excess, But It is Costly

When excess accumulates, the most common response is to push promotions. Barr cited that 69 percent rely on promotions as the number one tactic, even though it can erode margins. Liquidation was also used, and redistribution across warehouses was mentioned as an underused lever with real potential.

This matters because redistribution is one of the few tactics that preserves value. If one location is overstocked and another is understocked, transferring inventory can prevent a new purchase order and reduce waste. The barrier is often visibility. Without dashboards or clear reporting, teams do not know where excess sits or where demand is strongest.

Service Levels Remain a Defining Performance Marker

Despite volatility, many SMBs maintain strong service levels. Barr shared that 46 percent maintain service levels above 90 percent. Others operate in a broad middle band, and 14 percent do not even know their service levels, a visibility gap that increases risk.

The implication for supply chain planning is clear. High service levels require better demand forecasting, better safety stock logic, and faster response to supplier variability. If a company cannot measure service level, it cannot manage it, and it cannot connect inventory decisions to customer outcomes.

AI-Driven Planning is Accelerating Because It Reduces Friction, Not Because It is Trendy

Barr described AI adoption in SMB supply chain planning as cautious but accelerating. The tone was not hype. The focus was on practical outcomes: time saved, fewer manual steps, and clearer next actions.

SMB AI Investment is Rising, But Uncertainty Remains

He shared benchmark findings showing that 48 percent reported investing in AI, more than doubling from the prior year. A portion of respondents were unsure whether their business had invested in AI, which points to internal visibility and education gaps. This is common in SMB environments where tools may be adopted inside functions without broad alignment.

Looking forward, nearly half plan to invest within the next year, and the share with no plans declined. The pattern suggests momentum and a gradual reduction in uncertainty.

The Top AI Concerns are Data Security and Inconsistent Outputs

Barr was explicit about what holds teams back. The leading concern was data integrity and security risk. Inconsistent outputs were another issue because they erode trust across the organization. A small portion still did not understand how AI works, but that number declined, suggesting improving literacy.

This aligns with what most planning leaders already know. AI is only useful if the data foundation is trustworthy and if the outputs are consistent enough to guide decisions. When the stakes are inventory, cost, and customer service, trust is not optional.

AI Autonomy is Gaining Acceptance, With Human Oversight as the Bridge

One of the most interesting parts of the session was the discussion of autonomy. Barr described a pragmatic posture among SMBs: many would accept shared autonomy with human oversight, and a meaningful minority would fully embrace autonomy. A smaller group rejected it outright.

He framed this as trust built through repetition. When recommendations are consistently correct, teams begin to rely on them, and the role shifts from doing the work manually to supervising and validating what the system proposes.

Time Savings Translate Into Strategic Capacity, Not Headcount Cuts

In Q&A, Barr shared anecdotal evidence of 20 to 30 percent reduction in people time. He tied this to a familiar pain point: many organizations still forecast in spreadsheets, including large companies, and the effort can take days. With better software support, the same work can compress into a fraction of the time.

He also emphasized what he is not seeing: widespread reductions in force. Instead, the value is that teams can redirect effort toward supplier management, scenario planning, and growth support. The point is not fewer people. It is better use of skilled people.

Applying AI in the Planning Workflow: ERP Integration, Daily Item-Level Forecasting, and Scenario Planning

A recurring theme in the session was that modern supply chain planning cannot live in a vacuum. It must connect to ERP data, purchase history, and supplier performance, and then translate that data into decisions.

Integration Starts With ERP Data and Transactional History

Barr explained that the system ties into the ERP and ingests data such as purchase order history. He noted a practical constraint: meaningful demand forecasting and safety stock calculations require data history, typically two to three years of transactions. Without that, the planning engine has limited signal to work with.

This also surfaced in a question about organizations still using tools like QuickBooks rather than an ERP. Barr’s response was direct: planning tools like this generally require ERP-grade data. However, he described transitional scenarios where companies migrating to a full ERP may still be able to normalize and use historical data during the transition.

The Operational Core is Daily Forecasting and Real-Time Safety Stock

At a high level, Barr described the planning process as item-level forecasting done daily, with safety stock calculated continuously based on demand patterns and supplier performance. Supplier risk matters because lead time variability changes the inventory needed to protect service levels.

He also described the human behavior challenge: when teams panic, they revert to what they used to do. In uncertain markets, that instinct is common, but it often creates overbuying or inconsistent replenishment. The job of AI decision support is to make the logic visible and trustworthy enough that teams stay consistent.

Scenario Planning Becomes Essential Under Tariffs and Long Lead Times

Barr highlighted scenario planning as a key capability: what if you switch suppliers, pay more, but reduce lead time. What if lead times expand due to disruption. What if tariffs change the economics of a product line.

This came up in a question about long lead-time manufacturing and uncertain customer forecasts. Barr’s answer reinforced the value proposition: ERP systems may have planning features, but complex conditions like changing lead times and volatile supply performance often create gaps. In those gaps, organizations need better modeling and more actionable signals.

Turning Dashboards Into Action: The Hit List Approach

Barr shared an example of AI-driven inventory recommendations that prioritize action. Rather than forcing users to interpret many dashboards, the system can surface a list of the highest-impact actions, such as preventing a stockout, avoiding an overstocked purchase, or catching missed sales.

He described this as a second set of eyes. The core promise is not more data. It is clearer direction on what to do next.

He also emphasized a data governance principle: keeping data within the company’s environment rather than sending it to third-party servers, and using secure practices aligned with ISO 27001 standards. For many supply chain leaders, this governance posture is central to whether AI can be adopted at all.

Conclusion

Tariffs and supply chain disruptions are forcing SMBs to rethink how planning gets done. The organizations that remain resilient are not guessing their way through volatility. They are making deliberate trade-offs: absorbing costs to protect service levels, building strategic inventory buffers for high-impact items, and adopting alternative strategies like vendor-managed inventory to maintain flow. At the same time, the risk is real. Dead stock increases when hedging turns into overbuying, and working capital can vanish into inventory that does not move.

AI-driven supply chain planning is gaining momentum because it helps teams respond faster with less manual effort. The most practical applications are grounded in ERP data, supplier performance visibility, daily item-level forecasting, and scenario planning. Just as important, adoption is happening on SMB terms: cautious, focused on trust, and increasingly open to autonomy with oversight. For supply chain and operations leaders, the message is clear. The next planning cycle should prioritize visibility, actionability, and the ability to model uncertainty. If your team is still drowning in dashboards or losing days to spreadsheets, it may be time to evaluate how AI can turn planning into a repeatable, decision-ready process.

Contributors:

• Jefferson Barr, SVP Global Marketing at Netstock